evaluating electricity theft detectors in smart grid networks...“evaluating electricity theft...

18
Daisuke Mashima SEDN (Solutions for Electricity Distribution Networks) Group Fujitsu Laboratories of America Inc. Alvaro Cardenas University of Texas, Dallas Evaluating Electricity Theft Detectors in Smart Grid Networks

Upload: others

Post on 22-Oct-2020

2 views

Category:

Documents


0 download

TRANSCRIPT

  • Daisuke MashimaSEDN (Solutions for Electricity Distribution Networks) Group

    Fujitsu Laboratories of America Inc.

    Alvaro CardenasUniversity of Texas, Dallas

    Evaluating Electricity Theft

    Detectors in Smart Grid Networks

  • Advanced Metering Infrastructure (AMI)

    �Replacing old mechanical electricity meters with new digital meters

    �Enables frequent, periodic 2-way communication between utilities and homes

    Smart Meter

    Gateway Data Collection

    Metering Server

    GW

    Repeaters

  • Electricity Consumption Examples

    Weekly

    Daily

  • Electricity Theft under AMI

    Attacks will happen, but devices

    are deployed for 20~30 years.

    Strategy and tools for attack could

    be easily shared and distributed,

    e.g., through the Internet!

  • Taxonomy of Detection Mechanisms

    Detection of Electricity Theft

    Hardware

    Balance Meters

    Tamper Evident Seals

    SoftwareAnomaly Detection etc.

    Among software based detection, we focus on

    anomaly detection schemes because they do

    not require actual attack samples, which are

    hard to collect in practice.

  • Anomaly Detection Architecture in AMI

    Substation Houses

    MetersCollector

    Private Cloud

    Fib

    er-

    op

    tic

    ne

    two

    rk

    Router Router

    Smart Meters send consumption data

    frequently (e.g., every 15 minutes) to

    the utility

    Consumer 1

    Consumer n

    Electricity Usage

    Data Analytics,Anomaly Detection

    Meter DataRepository

    Storage

  • Our Contribution

    �Design anomaly-based electricity theft detectors using fine-grained electricity usage data reported by smart meters

    �Evaluate such electricity theft detectors

    �Instead of a traditional approach relying on real attack samples, propose new evaluation framework that uses “optimal” gain of attackers

    • I.e. find the worst-possible attack against each detector, and then calculate the cost (kWh stolen without being detected) of such an attack

  • Adversary Model

    Real Consumption Fake Meter Readings Utility

    Goal of attacker: Minimize Energy Bill:

    Goal of Attacker: Not being detected by classifier “C”:

    f(t) a(t)Compromised

    Smart Meter

  • � Take average of signal f(t) and report any average lower than a threshold as electricity theft

    � E.g. Select threshold as “2”

    � If daily-average of signal is lower than 2 report an alarm

    � Problem

    � Attacker, to maximize

    its gain, selects

    attack signal as

    constant a(t)=2

    Clearly a(t) looks

    “abnormal”, but it does

    NOT raise an alarm

    because the average of a(t) never went below 2!

    Detector using Simple Daily Average

    0

    1

    2

    3

    4

    5

    6

    7

    8

    3am 6am 9am 12pm 3pm 6pm 9pm 12am

    Normal Consumption 1 Attack

    f(t)

    a(t)

    Attacker’s

    gain

  • Other Electricity Theft Detectors

    �ARMA-GLR Detector

    �Use ARMA (Auto-Regressive Moving-Average) model to predict future consumption and evaluate the prediction error

    �EWMA (Exponentially-weighted Moving Average) / CUSUM (Cumulative SUM) Chart

    �Common techniques to continuously monitor process state (i.e Control Chart for QC)

    �LOF (Local Outlier Factor)

    �Clustering-based approach to identify outlying data points

  • Tradeoff Curves

    Y-axis: Cost of Undetected Attacks (can be extended to other fields)

    X-axis: False Positive Rate

    • Each detector is trained by using the last 28-day electricity consumption pattern.

    • Real AMI data (6 months of 15 minute reading-interval for 108 customers) is used.

  • Monetary Loss

    �Loss per customer

    �What if the attack propagates widely??

  • Effects of “Poisoning” Attacks

    � To incorporate changes in normal pattern over time (Concept Drift), detectors need to be re-trained periodically.

    � Attacker can use undetected attacks to poison training data

    “Valid” Electricity Consumption

    Undetected Attacks

    Time

    Re-train Detector toaccount for

    Concept Drift

  • Experimental Results of “Poisoning” Attacks

  • Detecting Poisoning Attacks

    �Identify concept drift trends helping an attacker

    �Continuously lower consumption over time.

    �Countermeasure: linear regression of trend

    �Slope of regression was not good discriminant

    �Determination coefficients worked!

    Honest Users Attackers Honest Users Attackers

    Slo

    pe o

    f R

    egre

    ssio

    n

    Dete

    rmin

    ation C

    oeff.

  • Ongoing Work

    �Use of cross correlation with other customers to detect attacks

    �Take “shape” of consumption curve into consideration?

    �Correlation with other factors? (Weather, temperature etc.)

    �Design and evaluate other detectors

    Distribution of cross covariance with other customers

  • Ongoing Work

    �Detect other types of anomalies

    �Apply LOF on consumption pattern of different customers on the same day

    �Outliers may be caused by a variety or reasons, such as meter failure etc.

    Typical patterns Outliers

  • Thank you very much.

    Contact:

    Daisuke Mashima

    [email protected]

    Fujitsu Laboratories of America Inc.

    1240 E. Arques Ave. M/S 345

    Sunnyvale, CA 94085

    �Reference:

    �“Evaluating Electricity Theft Detectors in Smart Grid Networks.” Daisuke Mashima and Alvaro Cardenas. In Proceedings of the 15th International Symposium on Research in Attacks, Intrusions and Defenses (RAID 2012), 2012.

    �Questions?